This paper investigates the estimation of uncertainty when evaluating the levels of human exposure to electromagnetic fields radiated from cellular base stations which aim to verify the exposure compliance with human protection guidelines. Uncertainty should be provided when publishing the results of an evaluation since measured physical quantities are not complete unless they are accompanied by some indication of their reliability. In this study, uncertainty sources are classified into four groups: physical parameters, mechanical constraints, postprocessing, and measurement equipment. Type A standard uncertainties are obtained from repeated observations by in situ measurement around cellular base stations, and they include uncertainties introduced by sources other than measurement equipment. Type B uncertainties are determined from the calibration data offered by measurement equipment manufacturers. The estimated overall uncertainty is +/-3.82 dB.